A FIBER STRESS-BASED PARAMETER FOR THERMOMECHANICAL FATIGUE LIFE PREDICTIONS OF TITANIUM MATRIX COMPOSITES

Titanium Matrix Composites (TMCs) are envisioned for use in the next g eneration of advanced aircraft and their engines. To ensure a smooth t ransition to industry, fatigue life prediction methodologies, which ca n account for random variations in mechanical and thermal loads, must be developed. To facilitate the development of such a model, fatigue t esting has been conducted at Georgia Tech. on [0/+/-45/90](s) and [90/ +/-45/0](s) laminates of SCS-6/Timetal 21S. The tests were done at tem peratures of 400, 500 and 650 degrees C, with hold times of 1, 10 and 100 s superimposed at the maximum stress. The purpose of the tests was to separate the effect of time-dependent deformation from the effect of environmental degradation. Using the results of these tests, and re sults generated at NASA-Lewis Research Center (LeRC) and the US Air Fo rce's Wright Laboratory, a model has been developed which is based on the stress in the load-carrying fibres. The stress is modified by an e ffective stress concentration factor that is due to matrix cracking an d a factor that includes the effect of hold times. It is a single term model that is intended for treating any variations in mechanical and thermal loads. Verification of this model is achieved by predicting fa tigue lives for specimens subjected to spectrum loads performed at NAS A-Lang]ey Research Center (LaRC) and vacuum tests completed at Georgia Tech. The model is compared to five methodologies previously develope d for life prediction, and is shown to have significantly better predi ctive power while reducing the number of empirical constants and curve fitting parameters necessary to collapse the data.